Upon culture of tissue biopsy samples from disease-free women, the majority of human mammary epithelial cells (HMEC) enter a proliferation barrier that is activated after several population doublings (1). A small subpopulation of human mammary epithelial cells, variant HMEC (vHMEC), have the ability to bypass this proliferation barrier and are typified by a silenced p16INK4a tumor suppressor gene, a sensitized stress response program (2) and the acquisition of a tremendous number of chromosomal abnormalities (1). Variant HMEC acquire several phenotypes critical to malignant transformation as a direct result of silenced p16INK4a. For example, in recent work we have shown that vHMEC contain an increased expression of the E2F1 transcription factor and several of its downstream targets such as the chromatin remodeling polycomb group (PcG) proteins, EZH2 and SUZ12. Upon loss of p16INK4a activity from parental HMEC cells, these PcG proteins are upregulated and localize to specific genomic sequences that are subsequently modified by DNA hypermethylation and silenced for transcription. We have demonstrated that loci targeted for DMA hypermethylation under these conditions include HOXA9, a transcription factor which controls cell fate and transcriptional programs important in mammary gland differentiation. We wish to use this powerful model system to investigate regulatory aspects of the HOXA9 DNA hypermethylation event as well as the identity of other genes that are targeted for DNA hypermethylation upon the silencing of p16INK4a in these cells. We will accomplish these goals through Specific Aims that seek to (1) determine the kinetics and extent of DNA hypermethylation at the HOXA9 regulatory region that takes place after the loss of p16INK4a activity in HMEC isolated from pre- and postmenopausal breast tissue and (2) determine if additional selected genomic sites that differ in DNA hypermethylation between HMEC and vHMEC are also dependent on the loss of p16INK4a activity. We will determine if the kinetics of DNA hypermethylation in this gene is similar to that measured at the HOXA9 locus. Finally, we wish to determine if the initiation of targeted HOXA9 DNA hypermethylation (brought about by loss of p16INK4a activity) can be phenocopied by the loss of other members of the pRb pathway. We hypothesize that the above-described properties of HMEC in vitro are critically relevant to their transformation processes in vivo and may provide clinically relevant biomarkers. [unreadable] [unreadable] [unreadable]